Dec. 30, 2024
12:44 p.m.
Thanks again for the speedy support.
The only (occasional) error message is that archiving failed ("archiving failed, re-queuing (mailing-list ...") but none of the above Mark mentioned. What is suspicious though is that uwsgi is at the top of the list of CPU slice consumers. Which may point at a problem with the REST-based communication between HK and MM.
Postgres only consumes roughly 5% of available CPU bandwidth but I'll review Greg's suggestions to see if these make any difference.